We developed a non-invasive and wearable system (1) that categorizes the characteristics of tremor and the daily activities and monitors how the frequency and amplitude of tremor are varied among daily activities through the analysis of continuous and long-term measurements (Phase-I), and (2) that utilizes peripheral stimulation techniques (i.e. electrical nerve stimulation, vibratory stimulation, and/or mechanical tremor-canceling perturbation) to modulate the tremor movements by optimizing the parameters of stimulations with closed-loop algorithm (Phase-II).